Abstract: In order to improve the moving speed of the transplanter manipulator without loss of robustness, the dc motor is selected as the driving motor, and the dual closed-loop speed control system model based on the current loop and the speed loop is designed according to the simplified transfer function model of the current feedback loop, the speed feedback loop, the filtering loop, the rectification loop and so on Introduction of BP neural network self-learning control strategy, including selecting the BP neural network PID controller to replace stages in PI speed loop, and using S function in MATLAB nested controller module way of double closed-loop system simulation model is built based on dc motor, the simulation results show that: compared to join the BP neural network algorithm is a double closed loop speed regulation system response curves before and after optimization, the optimized model of overshoot amount from 4.3% to 0 transition time reduced from above 2 S to 2 S, makes the motor speed control system stability and robustness, the whole row of manipulator start time shorter faster at the same time also can grab.